Emotion AI, Deauthorization, and Imposition

In this chapter, I draw on feminist studies, critical autism studies, and critical race and ethnic studies to examine emotion AI. Emotion AI is a set of narratives, technologies, and practices that impose normative emotion classifications. These impositions are structured by emotion AI’s foundational engagements with two contested and harmful theories of emotion: 1. the theory of basic emotions, which holds that certain human emotions are universal and universally expressed on the face, 2. the theory of mindblindness, which holds that autistic people lack the capacity to know their own emotions and the emotions of others. Given these theoretical engagements, I define emotion AI as a technology of deauthorization. I open by analyzing a narrative that highlights how emotion AI secures its authority by deauthorizing autistic people as knowledgeable emotional subjects. Through a discussion of emotion AI company Affectiva, I also consider the the racial colonial definition of the human that undergirds the theory of basic emotions, and thus most emotion AI systems. I conclude with a discussion of two artworks, Lauren Lee McCarthy and Kyle McDonald’s Vibe Check (2020) and Elisa Giardina-Papa’s Cleaning Emotional Data (2020), which open up political possibility in the disjuncture between the imposition of emotion classification and the irreducibility of emotional experience.